Les contraintes arithmétiques sont largement utilisées dans les langages formels comme les expressions, les grammaires d'arbres et les chemins réguliers. Ces contraintes sont utilisées dans les modèles de contenu des types (XML Schemas) pour imposer des bornes sur le nombre d'occurrences de noeuds. Dans les langages de requêtes (XPath, XQuery), ces contraintes permettent de sélectionner les noeuds ayant un nombre limité de noeuds accessibles par une expression de chemin donnée. Les types et chemins étendus avec les contraintes de comptage constituent le prolongement naturel de leurs homologues sans comptage déjà considérés comme des constructions fondamentales dans les langages de programmation et les systèmes de type pour XML. Un des défis majeurs en programmation XML consiste à développer des techniques automatisées permettant d'assurer statiquement un typage correct et des optimisations de programmes manipulant les données XML. À cette fin, il est nécessaire de résoudre certaines tâches de raisonnement qui impliquent des constructions telles que les types et les expressions XPath avec des contraintes de comptage. Dans un futur proche, les compilateurs de programmes XML devront résoudre des problèmes de base tels que le sous-typage afin de s'assurer au moment de la compilation qu'un programme ne pourra jamais générer de documents non valides à l'exécution. Cette thèse étudie les logiques capables d'exprimer des contraintes de comptage sur les structures d'arbres. Il a été montré récemment que le μ-calcul sur les graphes, lorsqu'il est étendu à des contraintes de comptage portant exclusivement sur les noeuds successeurs immédiats est indécidable. Dans cette thèse, nous montrons que, sur les arbres finis, la logique avec contraintes de comptage est décidable en temps exponentiel. En outre, cette logique fournit des opérateurs de comptage selon des chemins plus généraux. En effet, la logique peut exprimer des contraintes numériques sur le nombre de noeuds descendants ou même ascendants. Nous présentons également des traductions linéaires d'expressions XPath et de types XML comportant des contraintes de comptage dans la logique.

In the present work we analyze the algorithm that was introduced in [4]. The algorithm performs a translation of a conceptual schema to an XML schema expressed in the XML Schema language. We look for limitations of the algorithm and try to discover parameters that can be potentially used to influence its behavior. We propose solutions to the most serious limitations. Also, we introduce a concept of a translation profiling. The concept is based on a configuration that contains a set of parameters. We modify the algorithm to use the user requirements specified in the configuration. Thanks to the improvements, the new algorithm works with the concept of XML Namespaces, uses XML Schema designs and also, focuses on an elimination of redundancy. The elimination of redundancy in an output of the algorithm is an important part of this work and we create a formal model that helps us to solve this task.

In the presented work we study XML schema evolution, its types and impact on queries which are related on the particular schema. The thesis contains a review of existing approaches of this problem. The approach presented in this work shows a possible solution how to adapt related queries while schema evolves. The thesis contains a description of an algorithm which modies queries related to the evolved schema. Finally the work contains a number of exper- iments that verify proposal of the algorithms and show their advantages and disadvantages. 1

XML is a popular language for data exchange. However, many XML documents do not have their schema or their schema is outdated. This thesis continues on the field of automatic schema inferring for set of XML documents and focuses on Schematron schema inferring. Schematron is a language that validates XML documents with rules, it does not compare the document against a grammar like DTD, and XML Schema does. Because the field of Schematron schema generation is not so much explored, this thesis analyzes basic problems, suggests several approaches and describes their advantages and disadvantages.

In the data mining process, it is necessary to prepare the source dataset - for example, to select the cutting or grouping of continuous data attributes etc. and use the knowledge from the problem area. Such a preparation process can be guided by background (domain) knowledge obtained from experts. In the SEWEBAR project, we collect the knowledge from experts in a rich XML-based representation language, called BKEF, using a dedicated editor, and save into the database of our custom-tailored (Joomla!-based) CMS system. Data mining tools are then able to generate, from this dataset, mining models represented in the standardized PMML format. It is then necessary to map a particular column (attribute) from the dataset (in PMML) to a relevant 'metaattribute' of the BKEF representation. This specific type of schema mapping problem is addressed in my thesis in terms of algorithms for automatic suggestion of mapping of columns to metaattributes and from values of these columns to BKEF 'metafields'. Manual corrections of this mapping by the user are also supported. The implementation is based on the PHP language and then it was tested on datasets with information about courses taught in 5 universities in the U.S.A. from Illinois Semantic Integration Archive. On this datasets, the auto-mapping suggestion process archieved the precision about 70% and recall about 77% on unknown columns, but when mapping the previously user-mapped data (using implemented learning module), the recall is between 90% and 100%.

<p>Behovet av datautbyte mellan olika system, ökar ständigt. Därför byggs fler och fler system med möjligheten att utbyta och överföra data via XML. Då olika system lagrar data på olika sätt, måste de strukturella och innehållsmässiga skillnaderna hanteras innan data kan överföras.</p><p>Syftet med detta arbete är att undersöka hur dataöverföringar mellan XML-dokument i flera steg påverkas av strukturella eller innehållsmässiga skillnader. Arbetet visar i vilka fall data inte kan överföras korrekt. Studien utförs genom en experimentell metod. Experimenten genomförs med en applikation som utvecklats för ändamålet.</p><p>Resultatet från studien visar att problem kan uppstår då käll- och målstrukturen lagrar ett element eller attribut med olika datatyper, de övriga strukturmässiga skillnaderna genererar få problem. De innehållsmässiga skillnaderna medför att data inte kan identifieras då attributet saknas i käll- eller målstrukturen.</p>

During the development of federated IS that make use of non-homogenous databases and data sources, XML documents are often used for data exchange among the local subsystems, while their corresponding XML Schemas are generated using the standard CASE tools for local systems. External data schemes of those systems must be specified in a unified common model. An assumption is given that OBJECT PROPERTY (OP) model is being used for the semantic integration of the local non-homogenous subsystems. A graphical editor was developed that can be used to specify relation objects, their identifiers, complex and multi-valued object attributes. As OP model’s semantic expression capabilities can map those available in XML, additionally rules have been defined and implemented that can transform specific OP model structures into XML Schemas. Also algorithm is specified that can be used to extract tree-like structures from the model.Example transformations are performed that illustrate the process of generation of XML Schemas documents from sample OP models.

Behovet av datautbyte mellan olika system, ökar ständigt. Därför byggs fler och fler system med möjligheten att utbyta och överföra data via XML. Då olika system lagrar data på olika sätt, måste de strukturella och innehållsmässiga skillnaderna hanteras innan data kan överföras. Syftet med detta arbete är att undersöka hur dataöverföringar mellan XML-dokument i flera steg påverkas av strukturella eller innehållsmässiga skillnader. Arbetet visar i vilka fall data inte kan överföras korrekt. Studien utförs genom en experimentell metod. Experimenten genomförs med en applikation som utvecklats för ändamålet. Resultatet från studien visar att problem kan uppstår då käll- och målstrukturen lagrar ett element eller attribut med olika datatyper, de övriga strukturmässiga skillnaderna genererar få problem. De innehållsmässiga skillnaderna medför att data inte kan identifieras då attributet saknas i käll- eller målstrukturen.

Dieser Beitrag untersucht die Frage, wie komplexe Änderungen bei der Evolution von XML-Schemas automatisiert unterstützt werden können. Hierzu werden mögliche Änderungen an XML-Schemas durch Evolutionsoperatoren beschrieben, klassifiziert und beurteilt. Im Speziellen wird die Verschiebung (Move) von Elementen innerhalb von XML-Schemas analysiert. Die automatisierte Generierung und Ausführung von Transformationsregeln zur Migration von Instanzda-ten und die Beurteilung möglicher Informationsverluste während einer Transformation wird allgemein, wie auch anhand von Publikationsdaten untersucht.

Although XML is a widely used technology, the majority of real-world XML documents does not conform to any particular schema. To fill the gap, the research area of automatic schema inference from XML documents has emerged. This work refines and extends recent approaches to the automatic schema inference mainly by exploiting an obsolete schema in the inference process, designing new MDL measures and heuristic excluding of excentric data inputs. The work delivers a ready-to-use and easy-to-extend implementation integrated into the jInfer framework (developed as a software project). Experimental results are a part of the work.